Kids Are Using AI to Make Fake Nudes of Their Classmates. Most Schools Have No Plan for It.

Students are using cheap AI apps to make fake nude images of their classmates and pass them around school. Education Week reported on it August 6 as a back-to-school story. Most schools are opening the 2026-27 year with no policy, no reporting steps, and no staff training for it.

The apps are easy to find and the harm is severe. How well a school handles it depends on choices made before anything happens.

This is a plain explainer for Madison parents. What the apps are, what the law now says in Wisconsin, what to do first if it ever touches your kid, and what to ask your school while it’s still August.

What a “nudify” app actually is

Two kinds of app, same result. One takes a normal, clothed photo, like a team picture or an Instagram post, and uses AI to build a fake nude body under the real face. The other pastes a real face onto sexual images or video that already exist. Both produce an image of your kid that never happened and looks real.

Getting to one takes no skill. CBS News’s 60 Minutes counted over 100 of these sites. The biggest, Clothoff, was drawing over 3 million visits a month. The first images are free. After that, $2 to $40 apiece.

The sites claim to be 18+ and check nothing. A middle schooler needs a photo, a browser, and about a minute.

This is not a prank

The images are fake. What happens to the kid in them is not. In 2023, Francesca Mani was 14, a student at Westfield High School in New Jersey. She learned boys had run her Instagram photos, and her classmates’ photos, through Clothoff.

“I realized I should stop crying and be mad, because this is unacceptable,” she told 60 Minutes. The boy involved got a one-day suspension, her mother told 60 Minutes. Her case helped push the federal law that passed two years later.

How common is this? The Center for Democracy & Technology, a nonprofit that studies tech policy, surveyed US public high school students in mid-2024 . 15% said they had heard of an AI-made fake sexual image of someone at their own school. By CDT’s own extrapolation, that’s about 2.3 million students nationally. So the odds are real that a school near you has already had a case, whether or not any adult heard about it.

The same survey found girls are the target far more often. 51% of students who knew of cases said girls were depicted more often. Only 14% said boys were.

Two more numbers from that survey describe the whole problem. 71% of teachers said students caught sharing these images were expelled, suspended for days, or referred to police. Only 5% said their school helped the targeted student get the images taken down. Punishment exists. Help for the kid in the picture mostly doesn’t.

For now, assume that part falls to you.

The law changed this year

Three things moved, and they matter differently.

Federal: a 48-hour takedown right. The TAKE IT DOWN Act (S.146, signed May 2025 ) did two things. It made it a federal crime to publish intimate images of a real person without consent, and that covers AI-made fakes of real people. And it made platforms build a removal process.

That second part has had teeth since May 19, 2026. Per FTC guidance , a covered platform must remove a reported intimate image, real or AI-made, within 48 hours of a valid request from the victim. A parent can make that request when the victim is a minor. This is a duty on platforms, not on schools. But it hands you a lever parents didn’t have a year ago.

Illinois: the school law now names it. Illinois expanded its cyberbullying law to explicitly cover unauthorized AI-generated “digital replicas,” effective July 1, 2026 . One suburban principal, Daniel Krause of Willowbrook High School, told Education Week the change “will help to clarify any confusion and potential arguments that [cyberbullying laws] would not apply.” Illinois schools can now point at the statute.

Wisconsin: the crime is covered, the school side is not. Wisconsin’s school bullying law (Wis. Stat. 118.46 , on the books since 2010) requires every district to adopt and hand out a bullying policy. The law’s text says nothing about AI, deepfakes, or sexual images of any kind. It has not been updated the way Illinois’s law was.

Wisconsin criminal law did move. Since March 2024, under Wis. Stat. 948.125 , making or holding AI-made sexual images that appear to show a child is a felony. It sits on the same tier as real child sexual abuse images, even when no real body was in the photo. Prosecutors are using it. This year, Waukesha County charged a 20-year-old Illinois man who, prosecutors allege in the criminal complaint, used AI to make about 50 sexual images and videos of a real 15-year-old girl from Lisbon, WI. He has been charged, not convicted.

So in Wisconsin, the crime is defined. What your kid’s school must do about a case (who looks into it, how fast, what the targeted kid gets) is set by your district, not by state law. That gap is where families get stuck.

Two schools, two outcomes

In February 2024 at Beverly Vista Middle School in Beverly Hills, staff learned five 8th graders had used AI to make and share fake sexual images of 16 classmates. The district moved fast: it said the incident was contained within a day, and within about two weeks all five students were expelled , under agreements none of the families contested. Police opened their own investigation.

In Lafourche Parish, Louisiana, a 13-year-old girl and her friends reported to school staff — more than once, her family’s lawyers say — that a classmate had made and shared an AI fake nude of her. According to AP reporting , the images spread on Snapchat, which deletes messages after viewing, and the adults who looked said they couldn’t find them. Her family says the school failed to stop it. The district says it followed its protocols and that a one-sided story was presented.

When she saw the image being passed around again on the school bus, she hit the boy. She was expelled and sent to an alternative school for over ten weeks. What discipline, if any, the school gave the boy has never been made public.

After her family appealed, the school board let her return on probation, with the expulsion still on her record. The sheriff’s office later charged two students with 10 counts each of unlawful dissemination of AI-created images. Charges, not convictions — and the sheriff said the girl would not be charged.

The difference between the two districts wasn’t the size of the punishment. It was what happened between the first report and the punishment. The Louisiana girl did exactly what adults tell kids to do — she reported it, more than once, her family says. What she got, in the end, was an expulsion on her record.

Stanford’s Riana Pfefferkorn has reviewed more than 60 state and federal cases involving AI-made abuse images, and her research examines the same gap nationally: criminal law has moved much faster than school policies and procedures. Courts are still working out how to handle cases where both kids are minors. Which is why the school-level plan matters more than the size of the punishment.

Education Week found that the schools getting ahead of this share four habits. They name the reporting path out loud, tell students never to save or forward the images, offer counseling to targeted kids, and hold a “reentry meeting” (a planned sit-down before a disciplined student returns to class) instead of leaning on punishment alone. As one policy expert, Kristin Woelfel, put it: “It’s not just like ‘You’ll get in trouble if you do this.’ It should also be like, ‘Here is who you report something like this to.’”

If it happens to your kid

Save everything before you do anything else. Screenshot the image, the account that posted it, the timestamps, and the app or site where it appeared. Deleting it feels like protecting your kid. It destroys the evidence that both the platform takedown and the police will need.

Report it to a named person at the school. Not a note to the front office. Ask directly who looks into this and when you’ll hear back. Some schools have already picked someone. At Buffalo Grove High School in Illinois, the principal told Education Week an associate principal is positioned to respond quickly when a report comes in. If your school can’t name a person, that tells you where things stand.

File the platform takedown. Report the image to the platform where it’s spreading and cite the TAKE IT DOWN Act. The 48-hour clock is federal law now, and as a parent you can file for your child. Platforms wade through huge volumes of child-safety reports (Roblox alone logged 65,000 in a year ), so a specific request that cites the law and includes your saved evidence beats a vague flag.

Call the police, and don’t agonize over whether it “counts.” In Wisconsin, an AI-made sexual image of a minor can be a felony matter all by itself. You don’t have to decide that. Report it and let them decide. The criminal process and the school’s process run on separate tracks, and the school sitting on its hands doesn’t mean the police will.

Expect to drive it yourself. Remember the 5% number. Very few schools help families get images taken down. Assume you’re the project manager until your school proves otherwise.

What to ask your school this August

You don’t need an incident to ask these. Open house works fine.

• Does the written policy name AI-made or deepfake images, or just say “cyberbullying”? Illinois principals rewrote theirs so nobody could argue the policy didn’t apply.
• Who exactly handles a report, and how soon does the family hear back?
• What does the targeted student get? Counseling, and help with platform takedowns?
• What happens when a disciplined student comes back to class?

Vague answers in August are cheap information. Vague answers in the middle of a crisis are the Louisiana story. When Madison’s bullying reports jumped 30% , what helped wasn’t tougher punishment. And when we looked at what actually protects kids online , the answer kept coming back to adults who built the system before the bad day.

The apps are not going away, and the first families through this in any district are the ones who find out the hard way whether a plan exists. The girl in Louisiana reported it more than once. Nobody had decided in advance what would happen when she did. Ask your school for its plan while it’s still August.